deep-research
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill mentions an external repository (
https://github.com/Socialpranker/claude-deep-research) which contains a catalog of search channels, statistical sources, and API definitions. This is presented as a reference library for the agent to consult, not a direct remote code execution vector. The repository serves as an 'upstream' source for maintaining API and statistical registries. - [COMMAND_EXECUTION]: The methodology involves launching sub-agents and dispatching searches. This is part of the intended high-level agent workflow for parallel research and is not a vulnerability in the skill's instructions.
- [DATA_EXFILTRATION]: While the skill interacts with APIs and search channels, its primary purpose is data collection for the user's research. It includes a phase (3.5) for auditing available API keys in the environment, which is a common pattern for identifying available tools and is not inherently malicious in this context.
- [PROMPT_INJECTION]: The skill instructions are focused on process management and methodological rigor. There are no attempts to override safety filters or hijack the agent's core persona.
Audit Metadata